Project summary: We propose to establish an integrated MR imaging and analysis platform to examine hypoxic-ischemic (HI) brain injury in a neonatal piglet model, and to develop novel MRI markers that characterizes the evolving cellular pathology during injury progression. While MRI has been used extensively in HI, image interpretation and predictive accuracy of the conventional MRI markers, such as T1 and T2-weighted MRI or diffusion MRI (dMRI), leave much to be desired. In this project, we will develop microstructural MRI markers using the diffusion-time (td) dependent dMRI, which potentially improves the sensitivity and specificity of identifying cellular injury in HI. td-dMRI will be achieved by measuring water diffusivity at varying td's, using an oscillating gradient spin-echo (OGSE) dMRI sequence, to determine the td-dependency, which reflects the cell morphology. In a mouse model of neonatal HI, we have demonstrated that td-dMRI is sensitive to small microstructural changes in cells and subcellular organelles during early injury, and such microstructural details are not accessible by conventional dMRI. Here we will use a clinically-relevant piglet model of whole-brain HI, which exhibits well-defined phenotypes of gray and white matter injury that corresponds to human full-term newborns with birth hypoxia. Development of MRI markers in this model and investigation of the neuropathological substrates of the new MRI markers will have a high clinical impact. Clinical translation of td- dMRI, however, is challenging due to the gradient system on clinical scanners that limits the attainable td and detectable microstructural resolution. We will develop novel OGSE sequences to address the gradient limitation and evaluate the clinical potentials of td-dMRI using the piglet model. The study will be performed on 3T human scanners, and therefore, the MRI techniques will be readily translatable to clinical realm. We hypothesize that td-dMRI is sensitive to acute swelling of neurons and organelles after HI, and that early dMRI measures are predictive of long-term neuropathologic and neurologic outcomes. In Aim 1, we will build a piglet MRI platform with multi-metric MRI markers, including volumetric measures, high-order dMRI (DTI, DKI, tractography), magnetic transfer imaging, along with td-dMRI measures. We will also establish atlases of the developing piglet brains and atlas-based image analysis to achieve automated quantification of the multi-metric MRI data. In Aim 2, we will investigate the utility of td-dMRI in detecting microstructural injury during acute and subacute HI (6hrs - 7days), and explore the correlations between the early MRI markers with cellular and subcellular organelle pathology. In Aim 3, we will perform multimetric MRI to follow the injury progression in piglets over 30 days of recovery after HI, and evaluate the relations between early neuronal injury and white matter injury in the connecting tracts, as well as the long-term functional outcome with neurobehavioral tests. The MRI markers developed in this study will potentially improve the diagnosis and prognosis in clinical neonatal HI, which may lead to accurate and noninvasive evaluations of adjuvant therapies in these neonates.